Unveiling soil degradation and desertification risk in the Mediterranean basin: a data mining analysis of the relationships between biophysical and socioeconomic factors in agro-forest landscapes

Abstract

Soil degradation and desertification processes in the Mediterranean basin reflect the interplay between environmental and socioeconomic drivers. An approach to evaluate comparatively the multiple relationships between biophysical variables and socioeconomic factors is illustrated in the present study using the data collected from 586 field sites located in five Mediterranean areas (Spain, Greece, Turkey, Tunisia and Morocco). A total of 47 variables were chosen to illustrate land-use, farm characteristics, population pressure, tourism development, rainfall regime, water availability, soil properties and vegetation cover, among others. A data mining approach incorporating non-parametric inference, principal component analysis and hierarchical clustering was developed to identify candidate syndromes of soil degradation and desertification risk. While field sites in the same study area showed a substantial similarity, the multivariate relationship among variables diverged among study areas. Data mining techniques proved to be a practical tool to identify spatial determinants of soil degradation and desertification risk. Our findings identify the contrasting spatial patterns for biophysical and socioeconomic variables, in turn associated with different responses to land degradation.